We have a second Stats Society talk this month, this time at UTS. Details are below. Cheers, Michael =================== 6:00pm - 6.30pm: Refreshments 6:30pm - 7.30pm: Lecture 7:30pm - 8:00pm: Dinner Venue: The University of Technology Sydney, Building 2, Level 4, Room 23, Room 2.4.23, Corner of Harris St. & Thomas St., Ultimo, NSW 2007. Professor John Carlin Murdoch Children’s Research Institute & The University of Melbourne Assessing the risk of a rare adverse outcome following rotavirus vaccination: a case study in biostatistical methods and collaborative engagement Rotavirus infection is the leading cause of gastroenteritis in young children: worldwide it is estimated to cause 200,000 deaths and 10 million episodes of severe diarrhoea annually. The development of vaccines for this infection suffered a major setback in 1999 when the first licensed vaccine was withdrawn after it was linked with a substantially elevated risk of intussusception, a rare acute bowel obstruction that can have fatal consequences if not treated promptly. Since the mid-2000s, two second-generation vaccines have been introduced to many national immunisation programs following large-scale clinical trials that recruited infants in sufficiently large numbers to rule out adverse event risks of similar magnitude to those found with the first vaccine. However, post-marketing surveillance studies from Brazil, Mexico and Australia suggested that there might be risks with both of the new vaccines, at much lower levels than with the first vaccine. Given that any level of incidence of adverse events following vaccination is of major concern, the Therapeutic Goods Administration (TGA) commissioned us to lead a national study to assess the risk of intussusception associated with the two vaccines in Australia. Close collaboration with clinical experts was essential to this study but ultimately the main challenges were the ascertainment of good data and careful statistical analysis using both a self-controlled case series method and the more familiar case-control method. The self-controlled method relies on conditional likelihood inference and is challenging to understand intuitively and so to communicate effectively. The talk will provide an account of this project and of some lessons that may be drawn for effective biostatistical practice. I will also give an overview of the Victorian Centre for Biostatistics, an NHMRC-funded Centre of Research Excellence that aims to train biostatisticians at doctoral and postdoctoral level within the framework of a methodological and translational pipeline, extending from the discovery and evaluation of new methods to collaboration and consultation on a wide range of health problems. Biography of Professor John Carlin John Carlin has a national and international reputation in biostatistics, the science of developing and applying statistical methods to problems in health and medical research. This field is increasingly recognised as fundamental to modern research because of rapidly increasing technological capacity to accumulate and manipulate complex numerical data, in the face of which deep understanding of the theory and application of statistical methods has become ever more important. After completing a PhD in Statistics at Harvard University, John began a career in medical and public health research. As Director of the Clinical Epidemiology and Biostatistics Unit at the MCRI and Royal Children’s Hospital, he has played a central role in developing one of Australia’s leading biostatistical centres. This Unit has developed a broad program of work, encompassing basic training in clinical research methods, collaborative contributions to a wide range of clinical and population health research, with a growing focus on paediatric clinical trials, and its own methodological research program. In addition to his role at MCRI, John has a professorial appointment in the Centre for Epidemiology and Biostatistics, School of Population & Global Health, University of Melbourne. His international standing in the core discipline of statistics is attested by co-authorship of an influential graduate-level textbook with colleagues from Harvard University (Bayesian Data Analysis, 2nd edition 2003). He has over 300 scientific publications across a wide range of topics in statistical methodology and in numerous substantive clinical and public health areas.